Small Flask example for a RAG workflow: embed sample docs, embed the question, retrieve the most relevant sources, and answer with Telnyx AI Inference.

Feedback welcome, especially on what production RAG patterns or vector-store examples would be useful next.